The present invention relates to a migration method of the online data in a computer or a host computer (collectively referred to as a “computer” herein below) as well as a storage system, and more particularly to a migration method of the contents of a storage device including data subject to be used in business processing, without interrupting almost all computing. The present invention relates still more particularly to a data migration method from a existing storage system to a new storage system.
When updating an existing storage system into a new one, migrating data stored in the existing storage system into the new storage system is needed, in order to continue to utilize the data used in the processing hitherto. In such a case so far, data stored in the existing storage system would be to be backed up to a tape drive and the like at first, then the data thus backed up would be restored (written back) to the new storage device.
The backing up and restoring method as have been described above may use the processor incorporated in the computer. In order to prevent the consistency of data from being lost during the backing up operation by the alteration of the data to be backed up, any business processing that may use the data to be backed up has to be halted during the backing up operation. The period of time of stopping the business processing may be in general in the order of one to a few days, depending largely on the amount of data to be backed up and restored.
In the business of a recent data center, which is characterized by non-stop service of 24 hours a day, stopping the business processing for data migration in such a storage system may be a condition too difficult to accept. There have been proposed some data migration methods for improving the condition.
For example, a remote data copy method is disclosed in U.S. Pat. No. 5,544,347 wherein first storage system connected to a computer may automatically copy data into second storage system by itself, without intervention of the computer. In this system, data copy operation is hidden from the computer system in order to effectuate data copy into the second storage system without interrupting any business processing in the computer.
When applying this method to the upgrade scheme of storage system, utility is limited because the existing, first storage system has to have the remote data copy function. Also, when the system is operating, the contents of the first storage system is altered by the computer all the time, so that the contents in the second storage system transferred by means of the remote data copy method may not synchronize with any alteration made. In addition, since the existing, first storage system is located in a connection path different from the, path connecting the new, second storage system to the computer; the computer may recognize these storage systems as separate volumes. For these reasons, in order to use the data restored on a volume of the new storage system instead of the data on the existing storage system, the business processing program running on the computer has to be updated so as to reassign the volume to be accessed, even if the contents stored are the same.
There has been disclosed an online data migration system in U.S. Pat. No. 5,680,640, which may solve those problems encountered in case of the remote data copy methods, including the necessity of incorporating the remote copy functionality into the existing first storage system, inability of complete synchronization between the first and second storage systems, needs for updating the program controlling the computer due to the different connection path to the new storage systems, and so on.
In the above patent disclosed, the word “online” indicates that the reference to and alteration of the data subject to be migrated is allowed during migration. The second storage system in this migration system comprises a unit capable to effectuate the data restoration while at the same time allowing the sending/receiving (transfer) of data subject to the processing in progress in the computer. Whereby the computer may recognize the existing, first storage system as if the first storage is still connected thereto. In this system, it should be noted that the connection path between the computer and the existing, first storage system is disconnected and then the second storage system is connected thereto instead. Although the momentary disconnection of the connection path may cause the system to halt, the system may be dealt as being online since the period of time of the disconnection is far and away shorter than the backing up and restoring method as have been described above.
This method implicitly assumes a direct connection between the computer and the storage, namely the peer-to-peer connection, as well as the same interface to the new storage as the existing storage, namely the same communication protocol, resulting in the limited utility of the system.